We hypothesize that autoimmune diseases progress through several (at least two) "checkpoints", an early inflammatory and late phase destructive disease. We believe it might be possible to design common "prevention" strategies to arrest disease progression in one or both of these phases by the following. (1) Autoantigen reactive CD4+ T cells transduced with retroviral vectors to express "regulatory proteins" may provide tissue specific immunotherapy through the expression of the regulatory proteins at the site of autoimmune inflammation. (2) FcR non-binding anti-CD3 and anti-TNF antibodies might block disease progression. We plan to study the non-obese diabetic (NOD) mouse, that spontaneously develops insulin-dependent diabetes mellitus (IDDM) and shows many of the characteristics of human IDDM, and collagen induced arthritis (CIA) in the DBA/l mouse model, which shares certain characteristics of human rheumatoid arthritis (RA), to attempt these "disease prevention" strategies. We will use the antigen specific properties of autoantigen reactive T cells to develop an adoptive immunotherapy protocol to study the potential of one (or combinations) of various "regulatory" proteins to prevent both of these diseases. We will study the "anti-inflammatory" cytokines, IL-10 and IL-4, an antagonist of a "pro-inflammatory" cytokine receptor, IL-12 p40, and compare local delivery by retroviral transduction of auto antigen specific T cells with SC fV constructs of anti-CD3 and anti-TNF antibodies, to systemic use of the parent antibodies, in these two animal models, singly and in combinations. Additionally, we plan to analyze potential mechanistic effects of anti-CD3 on antigen reactive T cells compared to co-stimulatory blockade by looking at the expression of a novel anergy specific gene, GRAIL.